Type I or insulin-Dependent Diabetes Mellitus (IDDM) arises in genetically predisposed individuals as consequence of immune mediated destruction of the pancreatic islet insulin secreting beta-cells. The onset of clinical symptoms of diabetes represents the end point of a chronic progressive decline in beta-cell function, and occurs when the majority of beta-cells have been lost. Since type 1 diabetes develops insidiously from an ongoing, immune-mediated destructive process, and it can be predicted with some degree of accuracy, then it is possible that intervention at a point in this long prodromal period would prevent the initiation or perpetuation of beta-cell destruction. A number of potential intervention strategies exist that may slow the course or prevent the development of type I diabetes, and/or preserve beta-cell function and thus stabilize individuals with new onset type 1 diabetes. The Prevent Diabetes Group (PDG) has been organized on the premise that significant advances can be made by a cooperative approach to the design and conduct of human clinical intervention trials both to prevent development of type I diabetes mellitus and to preserve beta-cell function in recent onset type I diabetes mellitus. The goal is to create a dynamic and flexible environment where investigators can work together to develop and implement research protocols. Each member of the group supports the concept of inter-institutional studies by pooling clinical case material and laboratory resources with other collaborating members, and of participating in study committees for the purpose of developing new ideas and analyzing or evaluating the results of studies related to the prevention and/or stabilization of human type I diabetes mellitus. For the purposes of these studies, the disease is divided into several stages of development (early, advanced, and late prediabetes; new onset diabetes; etc.). The purpose of dividing subjects into these disease stages is that different intervention strategies might be tested in different groups. For example, one might be willing to accept a greater risk of side effects in late disease, e.g. new onset type I diabetes, than in any of the stages of prediabetes. Likewise, one might accept greater risk in late prediabetes than in early or advanced prediabetes. Moreover, different strategies may seem more applicable in one category than another. The PDG is made up of member institutions (clinical centers) and investigators organized into Committees, and facilitated by an Operations Office, a Statistical Office, and several Reference Laboratories. This provides the basic stricture for an ongoing series of trials. The Reference laboratories will assure standardized, reliable methods are used. Predictive models will be compared. In addition, we propose to conduct and monitor two pilot and feasibility studies to test the potential of intervention with [1] periodic courses of intravenous insulin, with or without accompanying chronic subcutaneous insulin, and [2] gut antigen presentation by oral administration of potential autoantigens to induce immunological tolerance.